GitHub

Allotaxonometry for all

Boys 1895 vs 1930

Here is the rank of the babynames for 1895

JohnWilliamJamesGeorgeCharlesFrankJosephRobertHenryHarryEdwardThomasWalterArthurFredAlbertClarenceRoyLouisSamuelCharlieErnestWillieEarlRichardDavidCarlJoeOscarWill0K1K2K3K4K5K6K7K8K

Here is the ranking for boy babynames in the US for 1930

RobertJamesJohnWilliamRichardCharlesDonaldGeorgeJosephEdwardThomasPaulFrankJackDavidRaymondKennethHaroldWalterBillyEugeneHenryArthurAlbertRalphCarlJoeHarryWillieGerald0K10K20K30K40K50K60K

If you wanted to compare which babyname got more popular over time, how would you do it? You can say that Robert is more popular than John. Then what?

1 2 10 20 100 200 1k 2k 10kRank rforBoys 1895more →frequent← lessfrequent 1 2 10 20 100 200 1k 2k 10kRank rforBoys 1930more →frequent← lessfrequent

The diamond plot is an histogram, where the axis are the rank of the tokens

A perfect diagonal would be if all names were of the same rank for both years

As you get further away from the diagonal on the right, you get the most popular boy babynames for 1930

Finally we can put everything together, contrasting two different years in different ways

Divergence

What are divergences? What are they doing down there?